Research On The Fast Method Of Determining The Crossovers Of Satellite

The orbit crossovers is the intersection of the trajectories of the sub-satellite points of the satellite orbits at two different moments.Observation and measurement of orbital crossovers are of special significance for earth survey satellites.Compared with traditional ground observations,the use of crossovers observations can analyze the changes in the system errors of the ground observation satellites by calculating the satellite crossovers discrepancies in the absence of accurate information on known points on the ground.Therefore,the crossovers data It is widely used in the field of satellite geodesy and oceanography.If you want to make full use of the observation information of the crossovers,you first need to accurately and completely determine the location of the crossovers.Therefore,how to quickly calculate the satellite crossing point from a large amount of measurement data has become a key issue.This paper analyzes the advantages and disadvantages of the existing algorithms,and partially improves the four crossovers calculation methods.The main work and conclusions are as follows:(1)In view of the fact that the line segment method cannot be applied to the calculation of multi-satellite crossovers,the rectangular shrinkage algorithm takes a long time to calculate the multi-satellite crossovers due to the shrinkage coefficient,and segmented fitting method loses crossovers in some cases.This article improves the above three algorithms.Using Jason-2satellite and HY-2 satellite data to conduct experiments and compare the above method with the rapid rejection straddle experiment method,the results show that each algorithm can calculate all the crossovers.Under the same conditions,the rectangular shrinkage algorithm in the single-satellite crossover experiment only takes 2.5 seconds,the segment fitting method is7 seconds,and the line segment method and the rapid rejection straddle experiment method take about 30 minutes.In the multi-satellite crossover experiment,the segmented fitting method is the fastest,and the rectangular shrinkage algorithm calculates slowly because of the shrinkage coefficient.(2)In order to reduce the search time of crossovers,combined with the idea of multithreaded programming,a partitioned parallel intersection processing strategy is proposed,which divides the global data into multiple threads for simultaneous calculation,which improves the calculation speed.Through the comparison of 4-thread,8-thread,and 12-thread single-satellite crossover experiments,it can be seen that the speed of the parallel program with12-thread partition is the largest.The comparison of the single-satellite crossover experiment of serial and parallel programs shows that the speed of the line segment method is increased by14.86 times,the speed of the rapid rejection straddle experiment method is increased by 9.41 times,the speed of the rectangular shrinkage algorithm is increased by 0.72 times,and the speed of the segmented method The legal speed is increased by 1.92 times.The comparison of the multi-satellite crossover experiment of serial and parallel programs shows that the speed of the line segment method is increased by 11.14 times,the speed of the rapid rejection straddle experiment method is increased by 4.95 times,the speed of the segmented fitting method is increased by 0.37 times,and the speed of the rectangular shrinkage algorithm An increase of80.83 times.Among them,the partition parallel rectangular shrinkage algorithm is the fastest in the calculation of single satellite and multi-satellite crossovers.(3)Based on the calculation results of the crossovers,the inverse distance weighting algorithm is used to calculate the sea level at the crossovers.The spatial position matching shows that the sea level at the crossovers of the Jason-2 satellite and the HY-2 satellite fluctuates from-100 m to 80 m,and the discrepancy value is from-1m to 1m.Through the 10-minute time difference matching,the sea level discrepancy value of the crossovers can basically be maintained at the cm level.Conditional adjustment is used to adjust the non-conformance value of the crossovers,and the linear and periodic errors in the error are removed by the combination of linear function and Fourier function.After adjustment,the root mean square of the nonconformance value is reduced from ±11.39 cm to ±6.5556 cm.